Post-Translational Regulation of the RSF1 Chromatin Remodeler Under DNA Damage

Molecules and Cells

Minireview

Post-Translational Regulation of the RSF1 Chromatin Remodeler under DNA Damage

Sunwoo Min1,2,3,5, Yong Won Choi1,4,5, Hansol Yun1,2,3, Sujin Jo3, Jae-Hoon Ji3, and Hyeseong Cho1,2,3,*

1Department of Biochemistry, 2Department of Biomedical Sciences, 3Genomic Instability Research Center, 4Department of He- matology and Oncology, Ajou University School of Medicine, Suwon 16499, Korea, 5These authors contributed equally to this work. *Correspondence: [email protected] http://dx.doi.org/10.14348/molcells.2018.2244 www.molcells.org

Chromatin remodeling factors are involved in many cellular INTRODUCTION processes such as transcription, replication, and DNA damage response by regulating chromatin structure. As one of chro- In response to DNA damage, DNA damage checkpoint sig- matin remodeling factors, remodeling and spacing factor 1 naling pathways are activated to transduce the signals and (RSF1) is recruited at double strand break (DSB) sites and to arrest the cell cycle to allow the repair of the broken DNA. regulates ataxia telangiectasia mutated (ATM) -dependent Three PI3-kinases are the major transducers, namely ataxia- checkpoint pathway upon DNA damage for the efficient re- telangiectasia mutated (ATM) kinase, ATM and Rad3-related pair. RSF1 is overexpressed in a variety of cancers, but regula- (ATR) kinase, and DNA-dependent protein kinase (DNAPK). tion of RSF1 levels remains largely unknown. Here, we These kinases phosphorylate their substrates and recruit showed that protein levels of RSF1 chromatin remodeler are repair proteins. Two types of DNA repair pathways are well- temporally upregulated in response to different DNA damage known; homologous recombination (HR) and non-homologous agents without changing the RSF1 mRNA level. In the ab- end joining (NHEJ). HR occurs mostly from the S to G2 sence of SNF2h, a binding partner of RSF1, the RSF1 protein phase, whereas NHEJ occurs at the G1 phase. DNA repair level was significantly diminished. Intriguingly, the level of and checkpoint pathways are significant in cells as the unre- RSF1-3SA mutant lacking ATM-mediated phosphorylation paired DNA may cause genomic instability, which may lead sites significantly increased, and upregulation of RSF1 levels to the diseases in humans (Ciccia and Elledge, 2010; Jackson under DNA damage was not observed in cells overexpressing and Bartek, 2009). ATM kinase. Furthermore, failure in the regulation of RSF1 In order to maintain genomic stability, in human, DNA is level caused a significant reduction in DNA repair, whereas wrapping up the histones forming nucleosome, and chro- reconstitution of RSF1, but not of RSF1-3SA mutants, re- matin remodeling factors exist to regulate the chromatin stored DSB repair. Our findings reveal that temporal regula- structure. In general, chromatin remodeling factors regulate tion of RSF1 levels at its post-translational modification by the chromatin structure during transcription and differentia- SNF2h and ATM is essential for efficient DNA repair. tion in eukaryotes (Nair and Kumar, 2012). Recently, it has been reported that chromatin remodeling factors play an Keywords: ATM, DNA double-strand breaks and repair, pro- important role in response to DNA damage for providing tein stability, RSF1, SNF2h access to DNA repair proteins to the site of the broken DNA (Lans et al., 2012; Osley et al., 2007).

Received 9 October, 2017; revised 21 November, 2017; accepted 27 November, 2017; published online 31 January, 2018 eISSN: 0219-1032 The Korean Society for Molecular and Cellular Biology. All rights reserved. This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/3.0/.

Mol. Cells 2018; 41(2): 127-133 127

Temporal Regulation of RSF1 Level under DNA Damage Sunwoo Min et al.

There are four conserved families of chromatin remodeling at the indicated time points. siRNAs of RSF1 and SNF2h factors in humans; the switch/sucrose non-fermentable were transfected using a electroporator (Neon); the cells (SWI/SNF), the Imitation SWItch (ISWI) ISWI, the chromo- were treated with phleomycin after 72 h and harvested at domain-helicase DNA binding protein (CHD), and the INO80 each time point. For MG132 treatment, cells were treated families. Among these families, the ISWI family includes the with MG132 at 5 h before harvesting the cells. common ATPase SNF2h, which has different types of binding partners such as RSF1, Williams syndrome transcription Western blot analysis factor (WSTF), and ATP-dependent chromatin assembly Gradient acrylamide gels (4–20%) were used and the analy- factor 1 (ACF1). The RSF complex is an ISWI family members, sis was performed as the standard procedure using previous- and comprises RSF1 and SNF2h (Aydin et al., 2014; Nair and ly described reagents (Min et al., 2014). The primary anti- Kumar, 2012). Previously, published clinical research has bodies used were as follow: RSF1 (Abcam and Abnova), reported that RSF1 is overexpressed in various cancers, and γH2AX (Millipore), SNF2h (Upstate and Abcam), GAPDH, that its overexpression is correlated with poor prognosis for GFP (Santa Cruz), pATM, pATR, phospho-p53(S15), pChk2 overall survival (Maeda et al., 2011; Shih Ie et al., 2005; Tai (Cell Signaling), Flag (Sigma), and V5 (Invitrogen). et al., 2012; Zhang et al., 2017). Furthermore, overexpression of RSF1 activates the DNA damage signaling pathway Immunoprecipitation by activating the ATM-Chk2 kinases (Sheu et al., 2010). Re- Cells were prepared as previously described (Min et al., cently, overexpression of RSF1 showed a positive correlation 2014). Briefly, cells were harvested and lysed in NETN buffer with p53 levels, and this combined expression of RSF1 and and sonicated. The lysate was cleared by centrifugation at p53 was correlated to tumor size and tumor, node, metasta- 13,000 rpm and the collected supernatant was incubated sis (TNM) staging in breast cancer patients (Ren et al., 2014). with the primary antibodies at 4℃ for overnight. Next day, Meanwhile, RSF1 has been reported to facilitate DNA dam- Protein A sepharose (GE) was added and the lysate was age response (DDR) signaling and repair. Previously, other incubated for 2 h and washed four times with NETN buffer. groups and we have reported that RSF1 facilitates ATM- For calf intestinal phosphatase (CIP) treatment, CIP (NEB) dependent checkpoint signaling and double strand break was incubated with the beads for 30 min at 37℃ after (DSB) repair (Helfricht et al., 2013; Min et al., 2014; Pessina washing the immunoprecipitated beads. and Lowndes, 2014). Thus, the temporal regulation of RSF1 levels is important for cell homeostasis upon DNA damage Repair assays to prevent endogenous DNA damage in cells and to facili- For repair assay, DR-GFP and EJ-GFP cells were seeded in a tate DSB repair in response to DNA damage. In this paper, 12-well plate and transfected with siRNA using lipofec- we reveal that RSF1 level is temporally regulated upon DNA tamine RNAiMAX (Invitrogen). Twenty-four hours after siR- damage by the formation of the RSF complex with SNF2h NA transfection, the cells were transfected with FokI endo- and its phosphorylation via ATM kinase. In addition, we sug- nuclease and harvested 48 hours after transfection. The gest that the misregulation of RSF1 level causes impaired harvested cells were washed with PBS, and analyzed by fluo- DSB repair. rescence-activated cell sorting (FACS).

MATERIALS AND METHODS Statistical analysis Statistics and graphs were performed using GraphPad Prism Cell lines and drug treatment (version 5.0). Unpaired student’s t test was applied to com- MCF7, 293T, EJ, and U2OS cell lines were cultured in DMEM pare two individual groups, while one-way ANOVA was high glucose (HyClone) supplemented with 10% FBS. The applied to compare multiple groups. For one-way ANOVA, MCF7, U2OS, and EJ cells were treated with DNA damaging the post-hoc test (Tukey HSD) was applied to test the signifi- drugs, such as phleomycin (50 μg/ml), etoposide (10 nM) cance between multiple groups using SPSS. Asterisks indi- and methyl methanesulfonate (MMS; 0.02% and 0.04%). cates each p-values (*P < 0.05; **P < 0.01, ***P < 0.001). Irradiation was treated with 10 Gray. The cells were harvested at each time points with 1x sample buffer and were sub- RESULTS jected to Western blot analysis. RSF1 stability is regulated at the post-translational level Site-directed mutagenesis upon DNA damage The primers used and procedures were previously described Because RSF1 has been reported to contribute in DDR sig- (Min et al., 2014). naling and DSB repair, we examined RSF1 levels in response to DNA damage (Min et al., 2014). The effect of DNA dam- Transfection aging agents such as phleomycin that induces DSB was ex- Cells were harvested after transfecting Flag-ATM and RSF1- amined. Interestingly, RSF1 levels increased significantly at GFP using lipofectamine 2000 (Invitrogen) and treating the the early time point after DNA damage (Fig. 1A). We also cells with phleomycin for 2 h or irradiation (10 Gy). For the examined the effects of etoposide in the U2OS cell line and cycloheximide chase assay, cycloheximide treatment was found that the RSF1 level was upregulated on treatment performed for 36 h after transfection of wild-type and 3SA with etoposide (Fig. 1B). In addition to the drug treatment, mutant using lipofectamine 2000 and cells were harvested

128 Mol. Cells 2018; 41(2): 127-133

Temporal Regulation of RSF1 Level under DNA Damage Sunwoo Min et al.

A B C D

E F G H

Fig. 1. RSF1 level is upregulated in response to different DNA damaging reagents at post-translational level. (A-D) U2OS cells were directly harvested in the sample buffer after treatment with phleomycin (A), etoposide (B), γ-irradiation (C), and MMS (D), and were analyzed by Western blotting with the indicated antibodies. (E, F) EJ cells (E) and MCF7 cells (F) were treated with MMS, followed by Western blot analysis. (G) MCF7 cells were harvested after treatment with MMS and analyzed by western blotting every ten minutes. (H) Total RNA was isolated from U2OS after treatment with phleomycin, and RNA level of RSF1 was analyzed by real-time PCR and normalized by that of GAPDH.

DNA damage by irradiation also induced stabilization of regulated at a certain time point depending on the cell line RSF1 (Fig. 1C). Furthermore, we examined the effects of and the damaging sources. These results also indicate that MMS, which is an alkylating reagent inducing multiple single the upregulated RSF1 level requires a fine-tuning mecha- strand breaks and DSB, and found that RSF1 levels were the nism for maintenance of the optimal RSF1 level upon DNA same as those observed on treatment with other drugs (Fig. damage. 1D). Because U2OS cell line is derived from osteosarcoma, Next, we measured the expression level of RSF1 mRNA in we also examined the regulation of RSF1 levels in epithelial DDR to examine if the upregulated level was dependent on cell lines. We observed the upregulated RSF1 levels upon its transcriptional level. RSF1 mRNA level remained un- DNA damage in EJ and MCF7 cell lines (Figs. 1E and 1F). The changed 2 hr after treatment with phleomycin (Fig. 1H). data showed that RSF1 level was upregulated immediately Thus, this result indicates that RSF1 level is upregulated upon after treatment with the four different DNA damage- DNA damage through its post-translational regulation. inducing-drugs. In order to observe the precise regulation of RSF1 stability, we harvested cells every 10 min for 1h, and The binding partner of RSF1, SNF2h, is important for the analysis of the data revealed that the level of RSF1 was tem- regulation of its expression upon DNA damage porally regulated in a time-dependent manner (Fig. 1G). In general, chromatin remodeling factors exist in a complex, These data suggest that the level of RSF1 increased signifi- and the subunits comprising the complex stabilize each oth- cantly, and the upregulated RSF1 expression was down- er (Watanabe et al., 2014). SNF2h is the most well-known

A B C

Fig. 2. RSF1 upregulation is dependent on the formation of the RSF complex. (A) U2OS cells were transfected with siCtrl and siSNF2h and treated with MMS (0.02%), followed by Western blot analysis. (B) U2OS cells were treated with siCtrl, siRSF1, and siSNF2h. At 48 h after siRNA transfection, cells were treated with MG132 for 5 h and harvested for Western blot analysis. (C) Total RNA was isolated from U2OS cells transfected with siCtrl, siRSF1, and siSNF2h by treating with MG132 for 5 h.

Mol. Cells 2018; 41(2): 127-133 129

Temporal Regulation of RSF1 Level under DNA Damage Sunwoo Min et al.

binding partner of RSF1 and forms the RSF complex with eral phosphorylation sites and among these sites, three RSF1. We tested if the stability of RSF1 was dependent on phosphorylation sites are the conserved motif of ATM/ATR SNF2h and found that the absence of its binding partner substrates. Based on RSF1 mass spectrometry, we per- significantly reduced the level of RSF1 in the presence and formed the phosphatase treatment of immunoprecipitated absence of DNA damage (Fig. 2A). We next examined if this RSF1 and found that RSF1 was a highly phosphorylated pro- phenomenon was mediated by ubiquitin-dependent prote- tein without DNA damage (Supplementary Fig. 1A). Fur- olysis; we treated MG132 to block proteasome-dependent thermore, protein stability is mediated by post-translational degradation. Western blot analysis revealed that the level of modification such as rapid phosphorylation by kinases (Zhao RSF1 was slightly, but not fully, recovered after treatment et al., 2017). Thus, we next examined if ATM kinase also with MG132 in the absence of SNF2h (Fig. 2B). We also influenced the protein stability of RSF1. Next we examined checked RSF1 mRNA level in SNF2h-depleted cells and whether RSF1 phosphorylation by ATM regulated RSF1 pro- found that the reduced level of RSF1 was dependent on tein stability upon DNA damage. By generating 3SA mutant post-translational regulation (Fig. 2C). Thus, we conclude (S524A, S1226A, and S1325A), which is unable to be phos- that the formation of RSF complex is required for the protein phorylated by ATM, we found that 3SA mutant showed stability of RSF1 in both absence and presence of DNA dam- high levels of RSF1, compared to WT, even in the equal age. amount of mRNA level of WT and 3SA mutant (Fig. 3A; Supplementary Fig. 1B). Next, we examined the half-life of ATM-mediated phosphorylation of RSF1 negatively 3SA mutant by the cycloheximide chase assay. We treated regulates its level upon DNA damage. the cells with cycloheximide to block translation and checked Figure 1 showed that the level of RSF1 was upregulated the half-life of the mutants, compared to that of WT. We upon DNA damage, and a fine-tuning mechanism was re- found that half-life of 3SA mutant was much longer than quired for maintenance of the optimal RSF1 level within few that of WT (Figs. 3B and 3C). Meanwhile, when we overex- hours. Previous reports showed that RSF1 is the direct inter- pressed ATM kinase, the level of RSF1 did not increase in acting protein with ATM kinase, which is the major kinase in ATM overexpressed cells (Fig. 3D). These data suggest that the DDR signaling pathway, and is the substrate of ATM/ATR RSF1 stability is also regulated by ATM upon DNA damage. kinase (Beli et al., 2012; Matsuoka et al., 2007; Pessina and Therefore, we suggest that post-translational modification of Lowndes, 2014). In addition to previous studies, RSF1 mass RSF1 mediated by ATM after DNA damage regulated protein spectrometry by our group revealed that RSF1 harbors sev- stability of RSF1.

A B C

D E

Fig. 3. ATM kinase fine-tunes the upregulated level of RSF1 upon DNA damage. (A) U2OS cells were transfected with V5, RSF1-V5 (WT), and 3SA-V5. At 48 h after transfection, cells were harvested and further analyzed by Western blot. (B) U2OS cells were transfected with RSF1-V5 (WT) and 3SA-V5 and treated with cycloheximide. Cells were harvested at the indicated time points and analyzed by Western blot. (C) Quantitative analysis of the Western blot shown in (B). (D) Flag-ATM was overexpressed in U2OS cell line. At 48 h after transfection, cells were treated with phleomycin for 2 h and analyzed by Western blot with the indicated antibodies. (E) U2OS RSF1 KO cells were transfected with siCtrl and siSNF2h. One day after siRNA transfection, cells were transfected with V5, RSF1 WT-V5, and RSF1 3SA- V5 and treated with MMS at 48 h after transfection, followed by western blot analysis with the indicated antibodies.

130 Mol. Cells 2018; 41(2): 127-133

Temporal Regulation of RSF1 Level under DNA Damage Sunwoo Min et al.

Given our observations that the stabilization of RSF1 was the level of γH2AX was significantly increased in RSF1- accompanied with the presence of SNF2h and post- overexpressed cells, compared to that in control cells (Sup- translational modification by ATM, we tested the combined plementary Fig. 2B). To examine the efficiency of DSB repair effect on protein stability of RSF1. Reconstitution of RSF1 in RSF1-overexpressed cells, we again counted GFP positive 3SA mutant and RSF1 WT in the RSF1 KO cell line showed cells in DR-GFP and EJ-GFP cell lines. Overexpression of ex- diminished level of RSF1 in the absence of SNF2h upon DNA ogenous RSF1 in DR-GFP and EJ-GFP cells in a dose- damage (Fig. 3E). Taken together, we suggest that both the dependent manner revealed that RSF1 overexpression also formation of RSF complex and its post-translation modifica- impaired efficient DSB repair, which was similar to the results tion by ATM are essential for the stabilization of RSF1 in the obtained upon RSF1 depletion (Figs. 4C and 4D; Supple- presence of DNA damage. mentary Fig. 2C). Thus, these data show that failure in tight regulation of RSF1 levels induced endogenous DNA damage Regulation of RSF1 stability is important for DSB repair response and prevented efficient DSB repair. To investigate the biological function of temporal regulation of RSF1 stability, we used DR-GFP and EJ-GFP cell lines to DISCUSSION measure the efficiency of HR and NHEJ, respectively, after depletion of RSF1 by siRNA transfection. Each cell line pro- Protein homeostasis for maintaining protein levels is usually duces GFP protein via HR and NHEJ repair system, after DSB important for signal transduction, particularly in DDR. p53 is is introduced by FokI endonuclease (Gunn et al., 2011). As the most well-known tightly regulated protein in response to previously reported, RSF1 depletion reduces the level of HR DNA damage, and its post-transcriptional modification is and NHEJ efficiency (Helfricht et al., 2013; Min et al., 2014; highly important for its stability (Lakin and Jackson, 1999). Pessina and Lowndes, 2014). Reintroduction of WT RSF1 Therefore, under urgent stress such as DNA damage, the slightly recovered HR and NHEJ efficiency, whereas the ex- optimal level of each protein involved in DDR is critical for pression of RSF1 3SA mutant failed to recover HR and NHEJ the maintenance of genomic stability. efficiency upon DNA damage (Figs. 4A and 4B). Thus these In this study, we focused on the protein stability of RSF1, results reveal that RSF1 regulation is important for DSB re- the level of which is significantly important for cell viability in pair. cancer cells. Previous reports have identified a positive corre- In addition, to examine the outcome of misregulation of lation between the overexpression of RSF1 by amplification RSF1 stability upon DNA damage, we overexpressed RSF1- of the RSF1 gene and the poor prognosis in many patients GFP in a doxycycline inducible cell line. As previously report- with cancer (Maeda et al., 2011; Tai et al., 2012; Zhang et ed, the overexpression of RSF1 induces an increase in the al., 2017). In addition to clinical research, molecular studies level of γH2AX, which indicates the activated DDR (Supple- in many cancer cell lines also showed that RSF1 overexpres- mentary Fig. 2A) (Sheu et al., 2010). Even after irradiation, sion induces the endogenous DNA damage by activating the

A B Fig. 4. Regulation of RSF1 level is required for the efficient DSB repair. (A, B) EJ-GFP (A) and DR-GFP (B) cell lines were transfected with siUTR-RSF1, followed by transfection with FokI combined with RSF1-V5 (WT) or 3SA-V5. GFP positive cells were counted by FACS analysis. (C, D) EJ-GFP (C) and DR-GFP (D) cell lines were transfected with RSF1-V5 combined with FokI in a dose dependent manner. GFP positive cells were counted by FACS analysis. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.005, one-way ANOVA with Tukey HSD.

C D

Mol. Cells 2018; 41(2): 127-133 131

Temporal Regulation of RSF1 Level under DNA Damage Sunwoo Min et al.

ATM signaling pathway (Sheu et al., 2010). Our data also supported by National Research Foundation of Korea grants showed that RSF1 overexpression increased the levels of the funded by the Korea government (MSIP) (No. 2011- endogenous DNA damage signaling pathway and impaired 0030043:SRC). efficient repair upon DNA damage. These data reveal that the optimal RSF1 level is required for cell viability and ge- REFERENCES nome integrity. Moreover, our data showed that RSF1 level is dynamic up- Aydin, O.Z., Vermeulen, W., and Lans, H. (2014). ISWI chromatin on DNA damage and the upregulation of RSF1 stability is remodeling complexes in the DNA damage response. Cell Cycle 13, mediated by the formation of RSF complex with SNF2h. 3016-3025. RSF1 stability is significantly upregulated in response to DNA Beli, P., Lukashchuk, N., Wagner, S.A., Weinert, B.T., Olsen, J.V., damage, followed by downregulation of its stability as Baskcomb, L., Mann, M., Jackson, S.P., and Choudhary, C. (2012). Proteomic investigations reveal a role for RNA processing factor γH2AX is induced. Because the maintenance of upregulated THRAP3 in the DNA damage response. Mol. Cell 46, 212-225. RSF1 level may prevent efficient repair, a fine-tuning mechanism of RSF1 level within its optimal level is tightly regulated Ciccia, A., and Elledge, S.J. (2010) The DNA damage response: making it safe to play with knives. Mol. Cell 40, 179-204. in DDR. Although the mechanism of upregulated RSF1 needs to be further explored, the presence of SNF2h in RSF Gunn, A., Bennardo, N., Cheng, A., and Stark, J.M. (2011) Correct complex is critical to RSF1 stability. Importantly, it is known end use during end joining of multiple chromosomal double strand breaks is influenced by repair protein RAD50, DNA-dependent that the level of the subunits of the SWI/SNF complex, a protein kinase DNA-PKcs, and transcription context. J. Biol. Chem. chromatin remodeling factor, are interdependent on each 286, 42470-42482. other and the presence of these subunits is required for the Helfricht, A., Wiegant, W.W., Thijssen, P.E., Vertegaal, A.C., recruitment of ATPase, BRM, in SWI/SNF complex at DSB Luijsterburg, M.S., and van Attikum, H. (2013). Remodeling and sites (Watanabe et al., 2014). Likewise, in RSF complex, spacing factor 1 (RSF1) deposits centromere proteins at DNA double- RSF1 is required for SNF2h accumulation at DSB sites (data strand breaks to promote non-homologous end-joining. Cell Cycle 12, not shown). However, we cannot exclude the possibility that 3070-3082. another subunit in ISWI family is also required for RSF1 sta- Jackson, S.P., and Bartek, J. (2009). The DNA-damage response in bility or SNF2h accumulation at DSB sites. human biology and disease. Nature 461, 1071-1078. In addition to SNF2h, ATM kinase is another player in the Lakin, N.D., and Jackson, S.P. (1999). Regulation of p53 in response temporal regulation of RSF1 stability in response to DNA to DNA damage. Oncogene 18, 7644-7655. damage. Previous study screening for the ATM/ATR sub- Lans, H., Marteijn, J.A. and Vermeulen, W. (2012). ATP-dependent strates upon DNA damage showed that RSF1 is one of the chromatin remodeling in the DNA-damage response. Epigenetics putative targets of ATM/ATR kinases after irradiation (Beli et Chromatin 5, 4. al., 2012; Matsuoka et al., 2007). Mass spectrometry re- Maeda, D., Chen, X., Guan, B., Nakagawa, S., Yano, T., Taketani, Y., vealed that RSF1 has multiple phosphorylation sites. In our Fukayama, M., Wang, T.L. and Shih Ie, M. (2011). Rsf-1 (HBXAP) previous study, we generated 3SA mutants and showed that expression is associated with advanced stage and lymph node 3SA mutant also impaired ATM-dependent DDR signaling metastasis in ovarian clear cell carcinoma. Int. J. Gynecol. Pathol. 30, pathway (Min et al., 2014). In this study, we suggest anoth- 30-35. er possibility that 3SA mutant upon DNA damage impairs Matsuoka, S., Ballif, B.A., Smogorzewska, A., McDonald, E.R., 3rd, DSB repair because of high expression of RSF1 due to the Hurov, K.E., Luo, J., Bakalarski, C.E., Zhao, Z., Solimini, N., Lerenthal, misregulation of protein stability. As shown in Figs. 4C and Y., et al. (2007). ATM and ATR substrate analysis reveals extensive protein networks responsive to DNA damage. Science 316, 1160- 4D, RSF1 overexpression also impaired DSB repair. Thus, 1166. tight regulation of RSF1 upon DNA damage is required for efficient repair. Min, S., Jo, S., Lee, H.S., Chae, S., Lee, J.S., Ji, J.H. and Cho, H. (2014). ATM-dependent chromatin remodeler Rsf-1 facilitates DNA damage In conclusion, we demonstrated that post-translational checkpoints and homologous recombination repair. Cell Cycle 13, regulation of RSF1 is required for the efficient repair in re- 666-677. sponse to DNA damage by the formation of RSF complex Nair, S.S., and Kumar, R. (2012). Chromatin remodeling in cancer: a with SNF2h and by post-translational modification of RSF1 gateway to regulate gene transcription. Mol. Oncol. 6, 611-619. by ATM kinase. Thus, these results provide insights to explain Osley, M.A., Tsukuda, T., and Nickoloff, J.A. (2007) ATP-dependent the mechanism by which RSF1-overexpressed cancer cells chromatin remodeling factors and DNA damage repair. Mutat. Res. respond to DNA damage inducing reagents and how atypi- 618, 65-80. cal regulation of RSF1 leads to genetic instability in cancer Pessina, F., and Lowndes, N.F. (2014). The RSF1 histone-remodelling progression. factor facilitates DNA double-strand break repair by recruiting centromeric and Fanconi Anaemia proteins. PLoS Biol. 12, e1001856. Note: Supplementary information is available on the Mole- Ren, J., Chen, Q.C., Jin, F., Wu, H.Z., He, M., Zhao, L., Yu, Z.J., Yao, cules and Cells website (www.molcells.org). W.F., Mi, X.Y., Wang, E.H., et al. (2014) Overexpression of Rsf-1 correlates with pathological type, p53 status and survival in primary ACKNOWLEDGMENTS breast cancer. Int. J. Clin. Exp. Pathol. 7, 5595-5608. We thank the members of Hyeseong Cho laboratory for Sheu, J.J., Guan, B., Choi, J.H., Lin, A., Lee, C.H., Hsiao, Y.T., Wang, critical discussions and Ie-Ming Shin, Susan Janicki, Jeremy T.L., Tsai, F.J. and Shih Ie, M. (2010) Rsf-1, a chromatin remodeling Stark, for kindly providing valuable reagents. This work was protein, induces DNA damage and promotes genomic instability. J.

132 Mol. Cells 2018; 41(2): 127-133

Temporal Regulation of RSF1 Level under DNA Damage Sunwoo Min et al.

Biol. Chem. 285, 38260-38269. and Yasui, A. (2014). SWI/SNF factors required for cellular resistance to DNA damage include ARID1A and ARID1B and show Shih Ie, M., Sheu, J.J., Santillan, A., Nakayama, K., Yen, M.J., Bristow, interdependent protein stability. Cancer Res. 74, 2465-2475. R.E., Vang, R., Parmigiani, G., Kurman, R.J., Trope, C.G., et al. (2005). Amplification of a chromatin remodeling gene, Rsf-1/HBXAP, in Zhang, X., Fu, L., Xue, D., Zhang, X., Hao, F., Xie, L., He, J., Gai, J., Liu, ovarian carcinoma. Proc. Natl. Acad. Sci. USA 102, 14004-14009. Y., Xu, H., et al. (2017). Overexpression of Rsf-1 correlates with poor survival and promotes invasion in non-small cell lung cancer. Tai, H.C., Huang, H.Y., Lee, S.W., Lin, C.Y., Sheu, M.J., Chang, S.L., Virchows Arch. 470, 553-560. Wu, L.C., Shiue, Y.L., Wu, W.R., Lin, C.M., et al. (2012). Associations of Rsf-1 overexpression with poor therapeutic response and worse Zhao, D., Lu, X., Wang, G., Lan, Z., Liao, W., Li, J., Liang, X., Chen, survival in patients with nasopharyngeal carcinoma. J. Clin. Pathol. 65, J.R., Shah, S., Shang, X., et al. (2017). Synthetic essentiality of 248-253. chromatin remodelling factor CHD1 in PTEN-deficient cancer. Nature 542, 484-488. Watanabe, R., Ui, A., Kanno, S., Ogiwara, H., Nagase, T., Kohno, T.

Mol. Cells 2018; 41(2): 127-133 133